teeth are usually retained until delivery of the immediate Probing of Pockets Related to denture, and they are easily accessible for clinical the Attachment Level* measurements. Measurements were made on 116 teeth. Measurements The measurements were obtained at the mesial and by distal facial line angles of the teeth. Reference marks JOHN F. SIVERTSON, D.D.S., M.s.† were cut in the crowns of the experimental teeth with a 556 crosscut fissure bur. The bur was held at the line FREDERICK G. BURGETT, D.D.S., M.s.† angles parallel to the long axis of the and was moved toward the tooth until a slight groove was cut in the clinical . The grooves were used as reference IT HAS NEVER been established to what extent a thin marks for all subsequent measurements. periodontal probe penetrates the epithelial attachment All measurements were made in the direction of the during clinical examination of pocket depth. Estimates long axis of the teeth using a Hiatt§ periodontal probe on a conceptual basis vary from no penetration to partial graduated in 3-mm segments. Four probes were used and or even complete penetration extending to the connective each probe was numbered for identification. Measure­ tissue attachment.3,8,30 ments by one investigator (F.G.B.) were recorded from The fact that the coronal border of the connective (1) the cementoenamel junction to the bottom of the tissue attachment can be stained12 as a definite demarca­ pocket and (2) the cervical border of the coronal bur tion on extracted teeth makes it possible to use this groove to the bottom of the pocket. All measurements landmark for comparison of pre- and postextraction were rounded to the nearest millimeter; except that measurements related to the cemento-enamel junction. anything close to 0.5 mm was always rounded to the The purpose of the present study was to relate lower whole millimeter. The patients were not anesthe­ post-extraction measurements of the stainable coronal tized and the bottom of the pocket was determined by the border of the connective tissue attachment to pre-extrac- probe meeting definite resistance but without provoking tion measurements obtained by probing in periodontal pain. pockets. Preparation for Bench Measurements MATERIALS AND METHODS After extraction in The University of Michigan De­ The clinical attachment level as related to the cemen- partment of Oral Surgery, the experimental teeth were toenamel junction of teeth with periodontitis and slated rinsed gently in tap water to remove blood and debris. for extraction were measured by one investigator They were then immersed in 4% methylene blue dissolved (F.G.B.). After extraction, the teeth were stained and in 50% alcohol for 1 minute and again rinsed in tap water remeasured by another investigator (J.F.S.). The pre- for 30 seconds. extraction and post-extraction measurements were com­ pared for differences, and the differences were analyzed Method of Bench Measurement for statistical significance. The identical numbered probe used for the clinical Selection of Patients measurement for each patient was again used for the bench measurement by an investigator (J.F.S.) other Immediate denture patients at The University of than the one who made the clinical measurements Michigan were screened for evidence of periodontal (F.G.B.). The distances from (1) the cementoenamel destruction. Twenty patients (10 women and 10 men) junction to the most coronal extension of the stained scheduled for immediate denture prosthesis and with connective tissue attachment and (2) the cervical border evidence of periodontitis consented to participate in the of the coronal bur groove to the most coronal extension study. The age of the patients ranged from 30 to 73 years. of the connective tissue attachment were measured. The The average loss of attachment as measured from the measurements were recorded only when these landmarks cementoenamel junction was 4 mm with a range from 1 could be identified clearly. All measurements were made to 9 mm. in the direction of the long axis of the tooth and were Selection of Teeth rounded to the nearest millimeter; except that anything close to 0.5 mm was always rounded to the lower whole Maxillary and mandibular , cuspids, and bi­ millimeter. Subsequently, dividing calipers were used to cuspids were selected for the investigation because these repeat the above measurements. The distances defined by the calipers were converted to millimeters using a * Submitted in partial fulfillment of the requirement for the degree of Master of Science in Periodontics. † Assistant Professor, Department of Periodontics, University § Marquis Dental Mfg. Co., 2005 East Seventeenth Avenue, Denver, of Michigan School of Dentistry, Ann Arbor, Michigan 48104. Colo. 80206. 281 J. Periodontol. May, 1976 282 Sivertson, Burgett millimeter scale with a Vernier scale and were recorded the cervical border of the coronal bur groove to the to the tenth of a millimeter. bottom of the pocket with bench probe measurements from the cervical border of the coronal bur groove to the Calibration most coronal connective tissue attachment in maxillary The first 15 teeth were measured at least three separate teeth. times in a random order, both clinicallly and benchwise. 3. Comparison of clinical probe measurements from These data were then analyzed to determine the the cementoenamel junction to the bottom of the pocket reproducibility of the measurements. with bench probe measurements from the cemento­ The standard deviation, average deviation per score, enamel junction to the most coronal connective tissue and the percentage of measurements showing no change attachment in mandibular teeth. at all were determined. 4. Comparison of clinical probe measurements from The analysis of scorer error (Table I) showed that the the cervical border of the coronal bur groove to the clinical measurements (recorded by F.G.B.) had a stan­ bottom of the pocket with bench probe measurements dard deviation of ± 0.07 mm for measurements from the from the cervical border of the coronal bur groove to the cementoenamel junction and ± 0.08 mm for measure­ most coronal connective tissue attachment in mandibular ments from the coronal groove. The average deviation, teeth. which is based on the algebraic sum of all deviations for 5. Comparison of bench probe measurements from all cementoenamel junction measurements was ±0.11 the cementoenamel junction to the most coronal connec­ mm, and for all groove measurement was ± 0.10 mm. tive tissue attachment with bench caliper measurements Seventy-two percent of the measurements from the from the cementoenamel junction to the most coronal cementoenamel junction and 70% of the measurements connective tissue attachment in maxillary teeth. from the coronal groove showed no change. 6. Comparison of bench probe measurements from The bench measurements (By J.F.S.) showed a stan­ the cervical border of the coronal bur groove to the most dard deviation of ±0.12 mm for measurements from the coronal connective tissue attachment with bench caliper cementoenamel junction with the probe and a standard measurements from the cervical border of the coronal deviation of ± 0.04 mm for measurements from the bur groove to the most coronal connective tissue attach­ coronal groove with the probe. The measurements from ment in maxillary teeth. the cementoenamel junction and coronal groove deviated 7. Comparison of bench probe measurements from an average of ± 0.11 mm per score, and 85% of both the cementoenamel junction to the most coronal connec­ measurements showed no change. The bench measure­ tive tissue attachment with bench caliper measurements ments using the caliper showed a standard deviation of ± from the cementoenamel junction to the most coronal 0.04 mm for measurements from the cementoenamel connective tissue attachment in mandibular teeth. junction and a standard deviation of 0.05 mm for 8. Comparison of bench probe measurements from measurements from the coronal groove. Each measure­ the cervical border of the coronal bur groove to the most ment from the cementoenamel junction and coronal coronal connective tissue attachment with bench caliper groove deviated an average of ± 0.14 mm and ±0.15 measurements from the cervical border of the coronal mm from the mean respectively. bur groove to the most coronal connective tissue attach­ The results represent a very low scorer error and a high ment in mandibular teeth. degree of reproducibility. The analysis of variance compared the means deter­ The caliper measurements in units of 0.1 mm cannot mined for each of the four variables. The significance for be compared meaningfully in percentage of no change each comparison was determined from each F value. To with probe measurements in units of 1.0 mm. equalize all cells in the analysis of variance only patients with identical teeth were used in each comparison. RESULTS AND DATA ANALYSIS The pairwise t test utilized the pooled measurements The measurements from all of the experimental teeth for all teeth. Means were determined for: clinical probe were recorded and analyzed by an analysis of variance measurements from (1) the cementoenamel junction and and a pairwise test. The analysis of variance tested the (2) the coronal bur groove to the bottom of the pocket, significance of interplay between four variables: quad­ and the bench probe and caliper measurements from (1) rant, patient, technique of measurement (bench and the cementoenamel junction and (2) coronal bur groove clinical), and surface measured. These were analyzed to the coronal extension of the connective tissue attach­ separately for each of the following comparisons: ment. These means were then compared with mean 1. Comparison of clinical probe measurements from difference, standard deviation, t statistic, and signifi­ the cementoenamel junction to the bottom of the pocket cance in the following pairings: with bench probe measurements from the cemento­ 1. Clinical probe vs. bench probe measurements from enamel junction to the most coronal connective tissue the cementoenamel junction to the attachment level. attachment in maxillary teeth. 2. Clinical probe vs. bench caliper measurements from the cementoenamel junction to the attachment level. 2. Comparison of clinical probe measurements from Volume 47 Number 5 Probing of Pockets 283

3. Bench probe vs. bench caliper measurements from listed pairings and in Table II. To equalize the cells in the cementoenamel junction to the attachment level. each pairing of the analysis of variance, teeth were 4. Clinical probe vs. bench probe measurements from grouped from patients all having the same teeth. The the coronal groove to the attachment level. interaction of the quadrant variable and surface variable 5. Clinical probe vs. bench caliper measurements from with each other or the other two variables was not the coronal groove to the attachment level. pertinent. The influence of the interaction of the patients 6. Bench probe vs. bench caliper measurements from variable and technique of measurement variable was of the coronal groove to the attachment level. paramount importance. The differences between the An analysis of variance for all the teeth was also means of the grouped teeth were not significant for any computed testing the interplay of patient and the tech­ pairing at the 0.01 level. The F values for the patient- nique of measurement (clinical and bench). The F value technique interaction were not significant for any of the and significance were determined. pairings at the 0.01 level (Table II). Therefore, the The analysis of variance (Table II) was utilized to variability between patients as to its influence upon the determine the influence of the interaction of the four technique of measurement (clinical and bench) was variables and whether their influence was significant. The negligible. means were calculated per patient for all variables and Since patient variability was not a factor, all of the comparisons were made according to the previously teeth were pooled. Means were calculated for all teeth

TABLE I. Analysis of Scorer Error

Average Standard Deviation % of Measurement Deviation Per Score No Change

Clinical Probe CEJ - Attachment Level .07 am .11 mm 72

Clinical Probe Groove - Attachment Level .08 mm .10 mm 70

tench Probe CEJ - CTA .12 mm 11 mm 85

Bench Probe Groove - CTA .04 mm 11 mm 85

Bench Caliper CEJ - CTA .04 mm 14 mm NA

Bench Caliper Groove - CTA .05 mm 15 mm NA

CEJ - Cementoenamel Junction CTA - Connective Tissue Attachment Groove - Coronal Bur Groove NA - Not Applicable

TABLE II. Analysis of Variance for Grouped Teeth

Pairings Cell Means F Value For Patient Technique Significance Clinical Bench Interaction At .01

Clinical Probe vs. Bench Probe Measurements from the CEJ to A.L. in Maxillary Arch 3.448 3.229 4.18 N.S.

Clinical Probe vs. Bench Probe Measurements from the Groove to A.L. in Maxillary Arch 5.823 5.833 1. 34 N.S.

Clinical Probe vs. Bench Probe Measurements from the CEJ to A.L. in Mandibular Arch 5.489 5.854 2.10 N.S.

Clinical Probe vs. Bench Probe Measurements from the Groove to A.L. in Mandibular Arch 7.403 7.611 1.90 N.S.

Pairings Cell Means P Value For Patient Techniques Significance Probe

Bench Probe vs. Bench Caliper Measurements from the CEJ to CTA in Maxillary Arch 3.229 3.159 .83 N.S.

Bench Probe vs. Bench Caliper Measurements from the Groove to CTA in Maxillary Arch 5.833 5.734 1.95 N.S.

Bench Probe vs. Bench Caliper Measurements from the CEJ to CTA in Mandibular Arch 5.864 5.760 3.47 N.S.

Bench Probe vs. Bench Caliper Measurements from the Groove to CTA in Mandibular Arch 7.611 7.514 2.80 N.S.

A.L. - Attachment Level Groovi • - Conona l Bur Groove CEJ - Cementoenamel Junction N.8. - Not Significant CTA - Connective Tissue Attach]sen t J. Periodontol. 284 Sivertson, Burgett May, 1976

from each patient, and a pairwise t statistic was deter­ not be penetrated painlessly by routine probing. Gott­ mined (Table III). The means were paired to compare lieb also based his theory on histological examination clinical and bench measurements from the cementoe­ of decalcified specimens but with a visible primary namel junction and coronal bur groove using both the enamel cuticle. According to his theory, the gingival probe and caliper results. Also, the means were paired to sulcus ended where histologically the primary enamel compare bench measurements from the cementoenamel cuticle met the crevicular epithelium. Periodontal junction and coronal bur groove using both the probe and pockets, according to his concept, would develop be­ caliper. The mean difference between the pairings was tween the tooth surface and the epithelial attachment, determined and the standard deviation calculated. The t and a probe could not penetrate this attachment without statistic was determined to test the null hypothesis that using undue force. Orban19 concurred with Gottlieb's the difference between the means was not significant. The theory when he found it impossible to detach the t statistic for all pairings was not significant at the 0.01 epithelium from the enamel during probing without level. The null hypothesis that the difference between the tearing the epithelium or connective tissue. means is zero was proven. Gottlieb's theory of an organic epithelial attachment The final analysis of variance for all teeth (Table IV) went almost unchallenged until Waerhaug30 returned to showed the patient-technique of measurement (clinical the concept of a nonexistent epithelial attachment in and bench) to be not significant at the 0.01 level. 1952. Waerhaug was able to pass a steel blade, 0.05 mm x 1.0 mm, to the coronal level of the connective tissue DISCUSSION attachment in dogs using only 7 gm of force. Using the Whether a probe in routine probing of pocket depth same probe, 250 gm of force was required to penetrate will penetrate painlessly through the epithelial at­ the mucous membrane of the lip. Histological examina­ tachment to the connective tissue attachment or stop tion of the specimen did not demonstrate tearing of the somewhere closer to the coronal border of the epithelial epithelial cells from the tooth surface. attachment always has been a controversial issue. Prior Orban17 repeated Waerhaug's experiments but was to 1921 it was felt that the epithelial attachment did unable to insert a probe to the connective tissue attach­ not exist, and that clinical probing penetrated to the ment. Histological evidence showed that the steel blades connective tissue attachment.3, 27 This concept was stopped within the epithelial tissue. The important dis­ based primarily on histological examination of de­ tinction between Waerhaug's and Orban's experiments calcified speciments. In 1921 Gottlieb8 introduced his was that Waerhaug directed the tip of the blade along the theory of an organic epithelial attachment which could tooth surface whereas Orban did not.

TABLE III. Pairwise T-Statistics for All Teeth

Mean Standard T- Significance Measurement Difference Deviation Statistic At .01 Level

CEJ-AL Clinical Probe 3.944 .081 .479 .758 N.S. CEJ-CTA Bench Probe 3.863

CEJ-AL Clinical Probe 3.944 .079 .462 .762 N.S. CEJ-CTA Bench Caliper 3.865

CEJ-CTA Bench Probe 3.863 .002 .135 .081 N.S. CEJ-CTA Bench Caliper 3.865

Groove-AL Clinical Probe 6.960 .138 .424 1.450 N.S. Groove-CTA Bench Probe 6.822

Groove-AL Clinical Probe 6.960 .195 .430 2.024 N.S. Groove CTA Bench Caliper 6.765

Groove-CTA Bench Probe 6.822 .057 .104 2.463 N.S. Groove-CTA Bench Caliper 6.765

AL - Attachment Level CEJ - Cementoenamel Junction CTA - Connective Tissue Attachment Groove - Coronal Bur Groove N.S. - Not Significant

TABLE IV. Analysis of Variance for All Teeth Cell Means P Value for Measurement Clinical Bench Bench Patient-Technique Significance Probe Probe Caliper Interaction at .01 Level

CEJ-AL or CTA 3.700 4.001 3.967 .241 N.S.

Groove-AL or CTA 6.635 6.736 7.176 .812 N.S.

AL - Attachment Level CEJ - Cementoenamel Junction CTA - Connective Tissue Attachment Groove - Coronal Bur Groove N.S. - Not Significant Volume 47 Number 5 Probing of Pockets 285

Numerous other investigators attempted to determine not significant at the 0.01 level. The differences that were the limits of the gingival sulcus by various methods. recorded were due to numerous possible causes. Al­ Zander33 forced cellulose acetate strips into the gingival though the exact location at which the measurements sulcus of a young dog. Histological examination showed were taken was marked by the coronal groove, slight that the strips had reached the connective tissue attach­ differences in the angulation away from the long axis by ment. Weinreb31 forcefully placed cavit and self-curing either investigator could produce a different reading.22 acrylic into the gingival sulcus of monkeys and was This is especially true on teeth with angular periodontal unable to separate the epithelium from the enamel. He defects. Clinical probing on extremely convex teeth is also placed steel blades (0.05 mm x 1.0 mm) and plastic hampered by the tonus of the gingival margin which strips into the gingival sulcus. If light pressure was might prevent the probe from following the curvature of applied, the blades and strips did not penetrate the the tooth.2 The extremely convex teeth also could cause a epithelial attachment. If force was used, the blades and difference between reading from the bench probe and the strips stopped within the epithelium. No separation of the bench caliper. As the number of measurements increased epithelium from the tooth occurred. Here again the tip of these slight inaccuracies were negated. the blade or strip was not kept in contact with the tooth Pocket depth measured from the cementoenamel junc­ surface. Henning and Zander10 used carbon insufflation tion proved as accurate as a pocket depth measured from in determining the limits of the gingival sulcus. At 25 psi the coronal groove. Although the cementoenamel junc­ they produced various planes of separations with some tion may be difficult to locate in some instances, it is a particles reaching the connective tissue attachment. The reliable landmark. The numerous irregularities of the rationale for this technique was that gas would be able to cementoenamel junction appear to balance out as the follow the path of least resistance and be capable of number of measurements increase. The coronal groove following the curvature of the tooth. produced as much variation as did the cementoenamel In other studies not specifically designed to determine junction. the limits of the gingival sulcus pertinent findings have The use of the calipers converting the measurements to been reported. In an experiment studying scaling, Ramf­ tenths of millimeters showed measurements by the jord and Kiester22 reported that routine scaling will tear periodontal probe to be very accurate. The differences and split the epithelial attachment and extend to the between measurements taken with the bench probe and connective tissue attachment in humans. In a study of bench caliper were very small and not significant at the healing of mucoperiosteal flaps Kohler and Ramfjord15 0.01 level. Therefore, the use of a periodontal probe reported a high correlation between clinical and micro­ graduated in 3-mm segments as used by the investigators scopic measurements of sulcus depth in humans. Mea­ in this study is very accurate in determining attachment surements were made with a blunted No. 2 silver point levels. from a coronal notch to the bottom of the gingival sulcus This study does not relate histological sulcus depth or pocket clinically and again from the coronal notch to with clinical sulcus depth. The histological sulcus is the the connective tissue attachment microscopically after shallow groove between the tooth and normal gingiva, extraction. Using the same method Costich and extending from the free surface of the junctional epithe­ Ramfjord7,21 also showed a high correlation between lium to the gingival margin.24 The clinical sulcus depth as clinical and microscopic sulcus in two studies investigat­ determined in this study cannot be related to the ing healing after denudation. The good correlation histological sulcus depth since the probe easily penetrated between the clinical and microscopic measurements the junctional epithelium. suggest that the probe penetrated to the connective tissue attachment. SUMMARY A good deal of the variation in results of conflicting Twenty immediate denture patients with periodontitis studies can be explained by the false equation of histolog­ participated in a controlled study to determine the ical and clinical sulcus depth. Histologically, sulcus depth relationship between clinical probing of pocket depth and is limited by the coronal border of the junctional the connective tissue attachment. Maxillary and man­ epithelium.24 However, measurement of clinical sulcus dibular anterior teeth were probed clinically by one depth does not appear to be limited by the epithelial investigator to determine attachment levels. The mesial attachment. Histological sulcus depth cannot be directly and distal facial line angles were probed from the related to clinical sulcus depth.24 Pocket depth measured cementoenamel junction and from a coronal bur groove, clinically may approach the connective tissue to the clinically determined attachment level. One hun­ attachment.24 dred and sixteen teeth were measured. In the present study the results show that a thin The teeth were extracted, rinsed, and stained with 4% periodontal probe reaches the connective tissue attach­ methylene blue in 50% alcohol to demonstrate the ment in routine probing of pockets in patients with remaining connective tissue attachment. A second inves­ untreated periodontal disease. The differences between tigator using the same probe as the first measured the the clinical measurements and bench measurements were distance from the coronal bur groove and cemento- J. Periodontal. 286 Sivertson, Burgett May, 1976

enamel junction to the most coronal extension of the 11. Hopwell-Smith, A.: An Introduction to Dental Anat­ connective tissue attachment. These measurements were omy and Physiology. Philadelphia, Lee & Febiger, 1913. repeated by the second investigator using a dividing 12. Hutchins, D. W., and Parker, W. A.: A study of plaque distribution in the gingival crevice. Oral Surg 35: 585, 1973. caliper and a millimeter scale with a Vernier. 13. Kelly, George Peter: Relationships of radiographic bone The data were analyzed by an analysis of variance for height, pocket depth, and attachment level in a longitudinal grouped teeth, pairwise t statistic for all teeth, and an study of periodontal disease, viii ± 76 p typed thesis. Ann analysis of variance for all teeth. The results showed: the Arbor, The University of Michigan, 1973. influence of the interaction between the patients and 14. Kiistala, N., and Mustakallio, K. K.: Dermoepithelial separation with suction. Electronmicroscopic and histochemic technique of measurement, bench and clinical, was study of initial events of blistering on human skin. J Invest negligible for the grouped teeth and for all teeth. The Dermatol 48: 466, 1967. difference between the clinical and bench measurements 15. Kohler, C. A., and Ramfjord, S. P.: Healing of gingival was not significant for all the teeth as well. The null mucoperiosteal flaps. Oral Surg 13: 89, 1960. 16. Kronfeld, R.: The epithelial attachment and so-called hypothesis that the difference between the clinical mea­ Nasmyths membrane. J Am Dent Assoc 17: 1889, 1930. surement and bench measurement is zero was satisfied. 17. Nasmyth, A.: On the structure, physiology, and pathol­ ogy of the persistant capsular investments and of the tooth. Trans Roy Med Chir Soc (London) 22: 310, 1839. CONCLUSION 18. Orban, B., Bhatia, H., Kollar, J. A., and Wentz, F. M.: 1. In routine clinical probing of untreated periodon­ The epithelial attachment (the attached epithelial cuff). J Periodontol 27: 167, 1956. tal pockets a thin periodontal probe will penetrate to the 19. Orban, B., and Mueller, E.: The gingival crevice. / Natl coronal level of the connective tissue attachment which Dent Assoc 16: 1206, 1929. must be assumed to represent the base of the epithelial 20. Parfitt, G. J.: Measurements of the strength of the attachment. attached epithelial cuff. Abstract. J Dent Res 40: 705, 1961. 2. A probe with markings for 3-mm segments, used by 21. Ramfjord, S. P., and Costich, E. R.: Healing after exposure of periosteum on the alveolar process. J Periodontol a well-trained investigator, will provide highly accurate 39: 199, 1968. and reproducible measurements of connective tissue 22. Ramfjord, S. P., and Kiester, G.: The gingival sulcus attachment levels for untreated teeth with periodontal and the periodontal pocket immediately following scaling of disease. teeth. J Periodontol 25: 167, 1954. 23. Schroeder, H. E., and Munzel-Pedrazzoli, S.: Morpho­ metry analysis comparing junctional and oral epithelium of REFERENCES normal human gingiva. Helv Odontol Acta 14:53, 1970. 1. Baume, L. J.: The structure of the epithelial attachment 24. Schroeder, H. E., and Listgarten, M. A.: Fine structure revealed by phase-contract microscopy. / Periodontol 24: 99, of the developing epithelial attachment of human teeth. Mono­ 1953. graph in Developmental Biology, Vol 2, Basel, S. Karger, 1971. 2. Beardmore, H. D.: Tonus of marginal gingiva. J Peri­ 25. Schultz-Handt, S. D., Waerhaug, J., From, S. H., and odontol 34: 31, 1963. Attramadal, A.: On the nature of contact between the gingival 3. Black, G. V.: Special Dental Pathology. Chicago, Medi­ epithelium and the surface. Periodontics 1: 103, co-Dental Publications, 1915. 1963. 4. Bodecker, C. F., and Applebaum, E.: The clinical 26. Stallard, R. E., Diab, M. A., and Zander, H. A.: The importance of the gingival crevice. Dent Cosmos 76: 1127, attaching substance between enamel and epithelium. A product 1934. of the epithelial cells. J Periodontol 36: 130, 1965. 5. Bodecker, C. F., and Lefkowitz, W.: Gingival reattach­ 27. Stillman, P. R., and McCall, J. O.: A Textbook of ment to the enamel and its relation to operative dentistry. Dent Clinical Periodontia, New York, Macmillan, 1922. Cosmos 77: 1106, 1935. 28. Toller, J. R.: The organic continuity of the , the 6. Cohen, B.: A study of the periodontal epithelium. Br enamel, and the epithelial attachment in dogs. Br Dent J 67: Dent J 112: 55, 1952. 443, 1939. 7. Costich, E. R., and Ramfjord, S. P.: Healing after 29. Turner, J. G.: The histology of the periodontal sulcus. Br partial denudation of the alveolar process. J Periodontol 39: Med J 1: 998, 1927. 127, 1968. 30. Waerhaug, J.: The gingival pocket. Odontol Tidsk 60: 8. Gottlieb, B.: What is a normal Pocket? J Am Dent (suppl.) 1, 1952. Assoc 13: 1747, 1926. 31. Weinreb, M. M.: The epithelial attachment. J Periodon­ 9. Gustafson, G., and Gustafson, A. G.: Structural and tol 31: 186,* 1960. Chemical Organization of Teeth. Vol. II, p. 75, New York, 32. Weiss, L., and Neiders, M. E.: The biophysical approach Academic Press, 1967. to epithelial cell interaction with teeth. Adv Oral Biol 4: 179, 10. Henning, F. R., and Zander, H.: A method for studying 1970. limits of gingival crevice and relative strength of epithelial 33. Zander, H. A.: A method for studying the "epithelial attachment. J Dent Res 42: 653. 1963. attachment." J Dent Res 35: 308. 1956.